1. Field of the Invention
The present invention relates to telecommunication networks, and more specifically to a method and apparatus for reducing overhead when using loopback cells for fault detection of bi-directional virtual circuits.
2. Related Art
Bi-directional virtual circuits are often provided to enable data transfers between two systems in a network. A bi-directional virtual circuit refers to a virtual circuit in which data traverses the same set of intermediate devices (forming the path) in both the directions. As an example, a virtual circuit provides communication between two end systems (e.g., edge routers or other devices interfacing directly with an ATM switch) in asynchronous transfer mode (ATM) networks, and the same intermediate switches provide the path between the two end systems in both directions.
Loopback cells are often used to detect faults in the path of virtual circuits. In a typical scenario, each end system (“source end system”) sends a loopback cell (“loopback command”) directed to the other end system (i.e., at the other end of the virtual circuit), and the other end system sends back the cell (“loopback acknowledgment”) with potentially some acknowledgment information incorporated within the cell. The source end system determines the status of the path (including intermediate devices and connecting transmission lines) depending on whether the loopback response is received or not.
OAM (operation, administration and maintenance) cells (both commands and acknowledgment/responses) used in the context of ATM networks are examples of the loopback cells. Only some of the details of OAM are described here for conciseness. For further details on OAM, the reader is referred to ITU_T Recommendation I.610 entitled, “Series I: Integrated Services Digital Network—Maintenance principles B_ISDN operation and maintenance principles and functions”, which is incorporated in its entirety herewith. At least in OAM related implementations, both end systems (at either end of a virtual circuit) independently check the status of the PVC.
One problem with such implementations is that the overhead on the components in the network due to resulting traffic (i.e., command and acknowledgment loopback cells) may be unacceptably high. The overhead may include use of bandwidth (which may otherwise be available for other applications) on the communication links and also buffering/processing overhead on all the devices in the virtual circuit path. Such overhead is particularly high on end systems which may need to process the cells more than other devices. The overhead may be unacceptable at least in that the corresponding end system implementations may not scale to large networks (employing a lot of end systems).
Accordingly, what is needed is a method and apparatus which reduces fault management related traffic (cells) when using loopback cells for fault detection of bi-directional virtual circuits.